CAREER: Characterizing and Optimizing Control in Neural Interfaces

University Of Washington, Seattle WA

Investigators

Abstract

This Faculty Early Career Development Program (CAREER) award supports research that aims to expand our understanding of human-technology interaction through Brain-Computer Interfaces (BCI). This award will support foundational research into how the nervous system responds to the use of neural interfaces, paving the way for the creation of advanced computer algorithms that can adapt to the user's nervous system. As BCI create intricate connections between the nervous system and technology by measuring biological signals from individuals and translating them into commands for devices, and they hold great potential for treating neurological disorders. Neural interface technologies have the potential to revolutionize the field of rehabilitation by allowing the nervous system to control novel devices that can offer new hope and possibilities for regaining control and independence despite physical limitations. However, developing computer algorithms that can effectively interact with the human nervous system remains a challenge. The interdisciplinary nature of this research will draw on the PI’s expertise in neuroscience, control theory, and neural engineering. Additionally, this award will support the creation of new outreach programs and integrate the findings into engineering courses while encouraging participation from underrepresented groups in engineering. Closed-loop interactions between a user and the device in a neural interface open opportunities to leverage nervous system plasticity to improve performance and shape user behavior for rehabilitation. Achieving this goal requires scientific insights into how nervous systems interact with devices and new computational frameworks to jointly consider the device, the nervous system, and their interactions. This project will identify principles of how users learn to control sensorimotor neural interfaces and use these insights to improve computational methods for closed-loop neural interfaces. The PI's team will perform experiments using two types of neural interfaces—muscle interfaces in humans and brain interfaces in non-human primates to understand computations performed by the nervous system when learning to control an interface and whether properties of the device influence these computations. The PI will quantify neural computations using a control theoretic framework that can measure users' predictive models of the device. Insights from these experiments will be used to improve user models, which will, in turn, be used to design new interface algorithms that will be experimentally validated against additional muscle interface experiments. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.